Salmon RNA-seq quantification for repeated genome

Question

[jordi.planells]

Hi all!

I have several questions regarding the usage of Salmon. I am trying to quantify the reads coming from the repeated genome (this is centromeres, transposons, simple repeats, etc.). I wonder whether using Salmon is a feasible/good way of doing so. If it is, then a second question pops out; which is the following:

Some of the repeats are smaller than the default K-mer choice (31), therefore are excluded from the index. Can I lower the K-mer value even though my data has a fragment length of 75bp?

If you argue than this is not the most suitable approach, could you suggest a different way of achieving what I am currently trying to do?

Thanks all before hand!

Cheers,

Jordi

[Dave Carlson]

If you have a database of repeats for your organism, you might want consider using either SalmonTE or TETranscripts

[jordi.planells]

Is there, conceptually any downside for using the regular version?

[Dave Carlson]

Nothing obvious to me. It might be worth running salmonTE and salmon on the same dataset (using a set of retrotransposons as the reference) and see if the results differ substantively.